Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Update Kokoro recipe

This commit is contained in:
Katsuya Iida 2021-05-29 19:39:51 +09:00
parent c4a5a73f18
commit 88f3255962
2 changed files with 17 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
recipes/kokoro/tacotron2-DDC/run.sh
View File

@ -1,22 +1,23 @@
#!/bin/bash
# take the scripts's parent's directory to prefix all the output paths.
RUN_DIR="$( cd "$( dirname "${BASH_SOURCE[0]}" )" &> /dev/null && pwd )"
CORPUS=kokoro-speech-v1_1-tiny
echo $RUN_DIR
# download LJSpeech dataset
wget http://data.keithito.com/data/speech/LJSpeech-1.1.tar.bz2
# extract
tar -xjf LJSpeech-1.1.tar.bz2
if [ \! -d $RUN_DIR/$CORPUS ] ; then
echo "$RUN_DIR/$CORPUS doesn't exist."
echo "Follow the instruction of https://github.com/kaiidams/Kokoro-Speech-Dataset to make the corpus."
exit 1
fi
# create train-val splits
shuf LJSpeech-1.1/metadata.csv > LJSpeech-1.1/metadata_shuf.csv
head -n 12000 LJSpeech-1.1/metadata_shuf.csv > LJSpeech-1.1/metadata_train.csv
tail -n 1100 LJSpeech-1.1/metadata_shuf.csv > LJSpeech-1.1/metadata_val.csv
mv LJSpeech-1.1 $RUN_DIR/
rm LJSpeech-1.1.tar.bz2
shuf $RUN_DIR/$CORPUS/metadata.csv > $RUN_DIR/$CORPUS/metadata_shuf.csv
head -n 8000 $RUN_DIR/$CORPUS/metadata_shuf.csv > $RUN_DIR/$CORPUS/metadata_train.csv
tail -n 812 $RUN_DIR/$CORPUS/metadata_shuf.csv > $RUN_DIR/$CORPUS/metadata_val.csv
# compute dataset mean and variance for normalization
python TTS/bin/compute_statistics.py $RUN_DIR/tacotron2-DDC.json $RUN_DIR/scale_stats.npy --data_path $RUN_DIR/LJSpeech-1.1/wavs/
python TTS/bin/compute_statistics.py $RUN_DIR/tacotron2-DDC.json $RUN_DIR/scale_stats.npy --data_path $RUN_DIR/$CORPUS/wavs/
# training ....
# change the GPU id if needed
CUDA_VISIBLE_DEVICES="0" python TTS/bin/train_tacotron.py --config_path $RUN_DIR/tacotron2-DDC.json \
--coqpit.output_path $RUN_DIR \
--coqpit.datasets.0.path $RUN_DIR/LJSpeech-1.1/ \
--coqpit.audio.stats_path $RUN_DIR/scale_stats.npy \
--coqpit.output_path $RUN_DIR \
--coqpit.datasets.0.path $RUN_DIR/$CORPUS \
--coqpit.audio.stats_path $RUN_DIR/scale_stats.npy \
--coqpit.phoneme_cache_path $RUN_DIR/phoneme_cache \

6
recipes/kokoro/tacotron2-DDC/tacotron2-DDC.json
View File

@ -2,7 +2,7 @@
"datasets": [
{
"name": "kokoro",
"path": "./kokoro-speech-v1_1-tiny/",
"path": "DEFINE THIS",
"meta_file_train": "metadata.csv",
"meta_file_val": null
}
@ -106,8 +106,8 @@
"max_seq_len": 153,
"compute_input_seq_cache": false,
"use_noise_augment": true,
"output_path": "./Models/Kokoro/",
"phoneme_cache_path": "./phoneme_cache/",
"output_path": "DEFINE THIS",
"phoneme_cache_path": "DEFINE THIS",
"use_phonemes": true,
"phoneme_language": "ja-jp",
"characters": {